All Sapphire 7950s (11196-19-20G) are not suitable for mining?

[deathkill]

no.. this has nothing to do with temperature...

i think reading through this thread that there is enough evidence that sapphire shipped a whole batch of faulty cards somewhere at the beginning of this year.

[LuckyLuke]

bought 2 cards. (11196-19-20G) tried to overclock them. both crashed. got new ones. tried to overclock from 925 to 950. crashed.
now I stopped overclocking and they are doing fine. arround ~550khash per card at 60-65C

1 question. If you overclock, do you need to put the boost option on or off? so overclock from 850 or from 925?

[sgggassfg]

no.. this has nothing to do with temperature...

i think reading through this thread that there is enough evidence that sapphire shipped a whole batch of faulty cards somewhere at the beginning of this year.

Yet, I can't find any source on the internet where overclockers or gamers are having the same problems as miners. The cards only seem to break down with mining...

Is that just because mining is more intensive or is there more too it?

[deathkill]

Yet, I can't find any source on the internet where overclockers or gamers are having the same problems as miners. The cards only seem to break down with mining...

Is that just because mining is more intensive or is there more too it?

true... i tried to look for it too.

somehow it is not a overclocking problem...

if i have the card without any overclocking and just set Intensity to for example 20 than i get crashes although according to the guy who wrote cgminer Intensity is not an overclocking setting i think:

https://bitcointalk.org/index.php?topic=87022.0
Quote from: ckolivas on June 12, 2012, 04:56:59 AM
---
Intensity is the amount of work in one go the GPU has to do before it can return its results. It needs to be large enough to keep the device fully utilised yet not so overloaded that the device becomes unresponsive. The faster the GPU, the higher the sweet spot is. Ram is of no use to mining whatsoever. They could even mine fine with only a few MB of ram.
---

[LuckyLuke]

is their a way to change the core voltage (mV) on msi afterburner for the sapphire 11196-19-20g

[LuckyLuke]

anyone?

[poohbah]

is their a way to change the core voltage (mV) on msi afterburner for the sapphire 11196-19-20g

No, but the Stilt bios for this card works well and it is at a lower than stock voltage. You can also edit the voltage with a bios editor such as VBE7.

[SaRmY]

no.. this has nothing to do with temperature...

i think reading through this thread that there is enough evidence that sapphire shipped a whole batch of faulty cards somewhere at the beginning of this year.

Yet, I can't find any source on the internet where overclockers or gamers are having the same problems as miners. The cards only seem to break down with mining...

Is that just because mining is more intensive or is there more too it?

My dead gpus had intensity 19.
In Gamers case, they don't play 24/7 and always at 99% gpu load, so i think that they haven't problems by these reason.

[sgggassfg]

I just burned yet another card today (5th card in less than 3 months). Again, no overclocking or anything.

This time I had set the fan speed to auto with a minimum of 60%, had 2 cards in the system of which the oldest one burned up (maybe 1 month old or so). The newer one still works fine.

The thing is... I had the fans to 70 for weeks and this was the first night I tried 60 fan speed. So maybe there's some truth in what some poster said earlier: the temperature sensor and automatic fan control on the cards is total shit and results in frying up a certain part of the card with no sensor when mining.


So my conclusion to this is: keep fan speed at 70% or higher and see if it helps...

[SaRmY]

I'm not sure about the auto fan can do this. In gpus with no conflictive SN like A1343000XXXXX and A1345000XXXXX i had my fans fixed at only 65% during 2 weeks (in holidays) about 3-4 hours each day this cards peaks temps up to 85-90ºC and no problem i have these gpus more than 9 months mining 24/7. Contra my dead cards had fixed fan at 80% with temps always under 75º C.

[LuckyLuke]

how come that when i overclock my 7950 it dies when i get the asus anti power surge triggered message. But when i dont overclock and i get the asus anti power surge message the cards are still working?

[SaRmY]

I think that I know the issue of all this... Yesterday to my cousin, died your first Sapphire 7950 (11196-19-20G) and like no, SN: A1351000286XX and today I see the card and all seem be ok, capacitors are ok but in the heatsinks of VMR's i saw an oil slick (and i remember that some of my gpus dead also had this oil slick) so I suspect seriously that any VRM exploited or burned and why Hynix VRM hasn't this problem, perhaps ELPIDA is a holy crap and Sapphire not be guilty of all.
So I recommended if you have any Sapphire (11196-19-20G) with SN starting for A1350 or A1351 not upload the memory core more than 1250 Mhz and if you have money, buy an accelero xtreme 7970 your VRM heatsinks are 100x better than Sapphire.

[jonnysomething]

I have 4 x Sapphire 7950s, they run well. One is shittier than the other three. Haven't been long term mining though.

What's everyone's hashrate? I get 620 on 3 and the other jumps between 590 and 620.

[LuckyLuke]

I think that I know the issue of all this... Yesterday to my cousin, died your first Sapphire 7950 (11196-19-20G) and like no, SN: A1351000286XX and today I see the card and all seem be ok, capacitors are ok but in the heatsinks of VMR's i saw an oil slick (and i remember that some of my gpus dead also had this oil slick) so I suspect seriously that any VRM exploited or burned and why Hynix VRM hasn't this problem, perhaps ELPIDA is a holy crap and Sapphire not be guilty of all.
So I recommended if you have any Sapphire (11196-19-20G) with SN starting for A1350 or A1351 not upload the memory core more than 1250 Mhz and if you have money, buy an accelero xtreme 7970 your VRM heatsinks are 100x better than Sapphire.

I got the A1351.

[SaRmY]

I think that I know the issue of all this... Yesterday to my cousin, died your first Sapphire 7950 (11196-19-20G) and like no, SN: A1351000286XX and today I see the card and all seem be ok, capacitors are ok but in the heatsinks of VMR's i saw an oil slick (and i remember that some of my gpus dead also had this oil slick) so I suspect seriously that any VRM exploited or burned and why Hynix VRM hasn't this problem, perhaps ELPIDA is a holy crap and Sapphire not be guilty of all.
So I recommended if you have any Sapphire (11196-19-20G) with SN starting for A1350 or A1351 not upload the memory core more than 1250 Mhz and if you have money, buy an accelero xtreme 7970 your VRM heatsinks are 100x better than Sapphire.

I got the A1351.

Afaik A135100028XXX and A135100029XXX have been conflictives for me. These gpus at 24/7 not lasted more than 2 months mining, in differents rigs, with different hardware and configuration but all the dead gpus had one thing in common the PN reference in all cards was 299-1E249-100SA and the SN starting always at A1350 and A1351.

My last 2 dead cards:

ASIC = 69,4% ---> 1,168v @ 1100 core - 1480 memory | 75%-fan-fixed | temps between 66º and 79º | SN: (A135000034XXX) (buy in France)
ASIC = 76,1% ---> 1,125v @ 1080 core - 1250 memory | 80%-fan-fixed | temps between 61º and 72º | SN: (A135100028XXX) (buy in Spain) 

[chipxtreme]

I bought four of these cards in January, one by one they dropped like flies, got one left. Ironically the three that have died had Hynix memory and the one remaining card has Elpida, and hashes much lower than the other 3 cards.

[EtBIM]

I bought four of these cards in January, one by one they dropped like flies, got one left. Ironically the three that have died had Hynix memory and the one remaining card has Elpida, and hashes much lower than the other 3 cards.

What temps did you run them at ?

Bought 3 of those 2 months ago, one died the first week, possibly because i ran them at a 80°C target, but the others survived probably because i keep them below or at 70°C now.

I can reach 620Kh/s with fans at 70%.

[sapstar]

I bought 5 of these cards 3 weeks ago and 1 died today. I keep them at 70°C. But the problem for me seems to be the fan. Two of my cards run at 90% speed. One of the fan became very slow today and then the card started producing hardware errors. I am still taking a replacement instead of money back as I can't find another card for this price.

Now I have spaced the cards a bit better and see fan rpms of 45%, 60%, 67% and 74% on the remaining 4 fans.

[tic-tac-totoro]

Sparks and purple smoke on 2 out of 4 cards now.

[LuckyLuke]

i think overclocking these cards make them die...